Re: [squid-users] cache_peer weighting

On Wed, 19 Dec 2007 09:33:26 -0700
Alex Rousskov <rousskov@measurement-factory.com> wrote:

> On Wed, 2007-12-19 at 23:31 +1300, Amos Jeffries wrote:
> > Squid is a little different in its peer requirements. The peering
> > needs to be anchored off the name= parameter or in the absence of
> > that the peer ip/fqdn given. In squid THAT must be unique for
> > several other indexes similar to this.
> > There is no requirement in squid that the IP/http-port combo be
> > unique because the IP/icp-port combo may be the difference.

Agreed; the odd thing is, there is no error handling in this case.
Perhaps it would be worth giving some visible messages within cache.log
too, if you have multiple cache_peers specified with the same IP,
taking part in a CARP array. Currently, it's only looking to make sure
there is a name= to make them distinct as far as I can tell. I've
modified my code so that it now hashes on names if "carp_hash_name on"
is in the squid.conf. It will fall back to the old method if it is
either off, missing, or if p->name is empty.

> What if Squid is not the only device doing CARP in a given
> environment? It may be important to allow the administrator to
> configure Squid hashing to match that of other devices so "extending"
> the hashing specs may require an option to turn the extensions off.

Agreed; as per this, the default option is for this extension to be
disabled by default, and must be explicitly enabled within the
squid.conf using the carp_hash_name config option.

http://www.squid-cache.org/bugs/show_bug.cgi?id=2153 is updated with
the latest patch.

-- 
Tony Dodd, Systems Administrator
Last.fm | http://www.last.fm
Karen House 1-11 Baches Street
London N1 6DL
check out my music taste at:
http://www.last.fm/user/hawkeviper